will it local
Launched May 19 2025 (~$299). LocalScore page (accelerator/3653) shows only Llama 3.2 1B Q4_K_M: 185 tok/s gen, LS 969, 1.59 s TTFT — no 8B Q4_K_M median published yet. TPU: GB206, 3840 CUDA / 120 Tensor cores, 28 Gbps effective GDDR7, 448 GB/s.
computed band 40.0-66.0 tok/s for 8B Q4_K_M (roofline, 448.0 GB/s VRAM)
Effective decode window: 0.46–0.76 of 448 GB/s nominal → ~206–340 GB/s effective (llama.cpp decode, Q4_K_M basis; arch window, community-calibrated)
→ 13b-q4 no 14.6 GB vs 8 GB usable — does not fit
→ 3b-q4 full 2.8 GB needed of 8 GB usable — headroom for context
→ 4b-q4 full 3.7 GB needed of 8 GB usable — headroom for context
→ 8b-q4 full 6.1 GB needed of 8 GB usable — headroom for context
no single-card bench published — band is computed (see notes)
Nominal bandwidth: 448.0 GB/s — see the effective decode window above
→ qwen3:0.6b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 0.6B · 0.6GB · needs ~1GB (4k ctx)
→ qwen3:1.7b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 1.7B · 1.4GB · needs ~1.8GB (4k ctx)
→ qwen3:4b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 4B · 2.6GB · needs ~3.2GB (4k ctx)
→ llama3.2:1b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 1B · 1.3GB · needs ~1.4GB (4k ctx)
→ llama3.2:3b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 3B · 2.0GB · needs ~2.4GB (4k ctx)
→ gemma3:1b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 1B · 1.0GB · needs ~1.1GB (4k ctx)
→ phi4:mini on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) mini 3.8B · 2.5GB · needs ~3GB (4k ctx)
→ moondream:2b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 2B · 1.7GB · needs ~3.7GB (4k ctx)
→ whisper:tiny on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) tiny · 0.1GB · needs ~2.1GB (4k ctx)
→ whisper on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) base · 0.3GB · needs ~2.3GB (4k ctx)
→ whisper:small on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) small · 0.9GB · needs ~2.9GB (4k ctx)
→ whisper:medium on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) medium · 3.1GB · needs ~5.1GB (4k ctx)
→ kokoro:82m on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 82M · 0.3GB · needs ~2.3GB (4k ctx)
→ gemma3:4b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 4B · 3.3GB · needs ~3.8GB (4k ctx)
→ deepseek-r1:7b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) distill 7B · 4.7GB · needs ~4.9GB (4k ctx)
→ mistral:7b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 7B · 4.1GB · needs ~4.6GB (4k ctx)
→ llava:7b on Raspberry Pi 5 + AI Kit (Hailo-8L) ($110) 7B · 4.7GB · needs ~6.7GB (4k ctx)
→ qwen3:8b on Tesla P100 (16GB, used) ($135) 8B · 5.2GB · needs ~5.8GB (4k ctx)
→ qwen2.5vl:7b on Tesla P100 (16GB, used) ($135) 7B · 5.6GB · needs ~5.8GB (4k ctx)
→ llama3.1:8b on Tesla P100 (16GB, used) ($135) 8B · 4.9GB · needs ~5.4GB (4k ctx)
→ deepseek-r1:8b on Tesla P100 (16GB, used) ($135) distill 8B · 4.9GB · needs ~5.4GB (4k ctx)
→ mistral-nemo:12b on Tesla P100 (16GB, used) ($135) 12B · 7.1GB · needs ~7.7GB (4k ctx)
→ qwen3:14b 14B — 9.9 GB vs 8 GB usable — partial CPU offload, expect large speed loss
→ gemma3:12b 12B — 9.6 GB vs 8 GB usable — partial CPU offload, expect large speed loss
→ deepseek-r1:14b distill 14B — 9.8 GB vs 8 GB usable — partial CPU offload, expect large speed loss
→ phi4:14b 14B — 9.9 GB vs 8 GB usable — partial CPU offload, expect large speed loss
→ whisper:large-v3 large-v3 — 8.2 of 8 GB usable — barely over; real with q8 KV cache (halves KV) or shorter context
→ sdxl SDXL base — 8.9 GB vs 8 GB usable — partial CPU offload, expect large speed loss
⚠ No 16GB variant exists — 8GB is the ceiling, so 13B+ Q4 models are off the table for local LLM work
⚠ PCIe x8 host interface (not x16) — measurable penalty on older PCIe 3.0 boards
⚠ Same 448 GB/s as the 5060 Ti but fewer SMs (3840 vs 4608 CUDA cores) — slower in compute-bound steps
⚠ LocalScore has no 8B Q4_K_M submission yet for this card (only Llama 3.2 1B) — bench_evidence left empty rather than fabricating an 8B figure
$480 — +8GB VRAM and 20% more SMs for a modest price bump — the natural tier-up
12GB / 672 GB/s / 6144 CUDA cores — breaks the 8GB wall that limits this card